Common FX loop for Matchless DC30

[ynor]

Greetings!

Once again I have another project: I've been missing my Matchless DC30 that I sold in 2009, regretting it ever since. Now I just received a 1994 SC30. One of the curiosities of this amp is the separate FX loop inserts for each channel, each returning into opposite sides of the PI. So using one FX processor (or pedal or something) for both channels needs some kind of workaround. Also, the FX loops are passive, and a solid state or tube-driven buffer unit ala D-lator/Kleinulator is needed. I would like to hear your thoughts on approaches for summing the two channels to one FX loop, applying a buffer unit for FXs. In my case, I do have a D-lator (from C-tone, modified with toroid PT), but I currently use this with my Dumble clone. But another D-lator with two inputs that sums the channels might be used? Other ideas? The two channels are out of phase.

DC30CH1a.gif

DC30 CH2a.gif

DC30 OUTa.gif

DC30 PWRa.gif

DC30schematic rev 1995.gif

[ynor]

I came across the Trinity TC-15 schematic, which is a 15 W clone of the DC30 with a common FX loop and added reverb that gave me an idea. Note that there is one error in this schematic: There is a node at R12, R13, R14 and R15, not jump.

TC15 kopi.gif

Concept: I want to use a single FX processor for both individual FX inserts for each channel of a Matchless DC-30 without modifying the amp. The solution I propose is to build an external summing box that interfaces with my existing C-lator (a Dumble-style tube buffer/recovery unit).

Summing the sends (in the passive external box):
Take the FX send from Channel 1 and Channel 2 (from the DC-30’s insert points) with stereo jacks into a passive external box.
Each goes through its own series resistor (e.g. 150k) into a summing node.
The summed signal is then sent via a mono jack to the C-lator for buffering and driving the FX processor.

Return / recovery (in the C-lator):
The FX return from the processor goes into the C-lator’s recovery stage, which has adjustable gain.
This allows me to compensate for any loss from the summing resistors and ensure that the DC-30’s phase inverter (PI) sees the same level it would have had without the loop engaged.

Switchable output impedance (back in the summing box):
The recovered signal from the C-lator is sent back to the summing box via a mono jack.
Inside the box, there is a switch that selects between:
Low-Z mode.
High-Z mode: insert a ~100k series resistor, so the PI sees a high source impedance similar to the case with bypassed fx loops (no insert cables).

Single-grid return (critical point):
The return signal is routed back to only one of the FX loops, i,e one PI grid, not both.
This asymmetry is what happens in a stock DC-30 (or Vox AC-30) when you play only one channel at a time.
Driving a single PI grid is a fundamental part of the Matchless sound, since the PI clips asymmetrically and contributes strongly to the amp’s signature crunch.

Any thoughts on this idea?

[martin manning]

Any thoughts on this idea?

I'd be reluctant to modify this amp, but I think I would try some small reversible modifications to see if mixing the two channels can be made to work satisfactorily. Something like this:

[pdf64]

... C-lator (a Dumble-style tube buffer/recovery unit). ...

Do you realise that the 250k send loop level control in Dumble type loops makes the preceding send buffer somewhat pointless?

I'm not saying they don't produce super awesome toanz, just that they aren't an example of a transparent fx loop, because the send impedance won't be low (and may be rather high, up to about 63k ohm), unless that send level control is turned up to full CW.
The loop send level seems to be labelled 'drive' on the Ceriarone version.

https://el34world.com/charts/Schematics ... e/Loop.pdf

https://ceriatone.com/wp-content/upload ... Layout.jpg

[lonote]

^^^

I had a couple of thoughts/questions after you made that same point in another thread a day or so ago.

Wouldn't the issue be dependent on what is plugged into the FX loop? A piece of rack gear would probably prefer a low impedance signal but pedals are usually getting a direct signal from the instrument at a similar high impedance. Not as big a deal..?

Your point also reminded me of Martin's tweaked Dumbleator schematic that replaces that 250K send pot with a 50K. Seems that should help some with the issue.

[ynor]

I'd be reluctant to modify this amp, but I think I would try some small reversible modifications to see if mixing the two channels can be made to work satisfactorily. Something like this:

Yes, I would not do any modding of this amp. It's a Sampson era all-original so it does have some collectors value.

Your suggestion for a mod looks quite similar to the suggestion posted in my second post?
I think this can be done without any mods to the amp itself: Each of the independent FX loops uses a stereo insert cable, with tip carrying send and ring carrying return to each of the grids of the phase inverter. If I make a cable with two TRS, one for each channel, join the tips with a 220k in series with each of the tips (summing the channels) to a mono jack, send that to the D-lator and return from the D-lator to ring in only one of the FX loops. Then both FX-loop sends go to the D-lator, and only one return goes into one side of the PI.

[ynor]

Do you realise that the 250k send loop level control in Dumble type loops makes the preceding send buffer somewhat pointless?

I'm not saying they don't produce super awesome toanz, just that they aren't an example of a transparent fx loop, because the send impedance won't be low (and may be rather high, up to about 63k ohm), unless that send level control is turned up to full CW.
The loop send level seems to be labelled 'drive' on the Ceriarone version.

Yes, the C-lator does color the tone quite. I'm thinking of just trying with the C-lator first just to see if summing the fx sends works, and maybe get another FX buffer that's more transparent later as the C-lator will be occupied in my Dumble clone. Maybe something like the Kleinulator or similar.

I plan to use my Quad Cortex in the FX loop.

[pdf64]

... Each of the independent FX loops uses a stereo insert cable, with tip carrying send and ring carrying return to each of the grids of the phase inverter. If I make a cable with two TRS, one for each channel, join the tips with a 220k in series with each of the tips (summing the channels) to a mono jack, send that to the D-lator and return from the D-lator to ring in only one of the FX loops. Then both FX-loop sends go to the D-lator, and only one return goes into one side of the PI.

For the LTP to work properly, the ring of one of the TRS plugs would need grounding.
Or just use a mono TS plug for one, TRS for the other.

[pdf64]

^^^
Wouldn't the issue be dependent on what is plugged into the FX loop? A piece of rack gear would probably prefer a low impedance signal but pedals are usually getting a direct signal from the instrument at a similar high impedance. Not as big a deal..?

Your point also reminded me of Martin's tweaked Dumbleator schematic that replaces that 250K send pot with a 50K. Seems that should help some with the issue.

Yes a 50k pot would be a significant improvement, in regard of loop transparency.

The send impedance should be low to avoid cable capacitance rolling off the high end, and to reduce the tendency for the cable to pick up hum from AC magnetic fields (eg from mains transformers) it passes through.

I suppose if the send was plugged into a rack unit with a typical studio medium input impedance eg 10k then that would mitigate for the 63k send output impedance, total effective impedance would then be 8k6.

Yes, a guitar's output impedance will be high, but the tonal balance of a regular electric guitar relies on a certain degree of cable capacitance, eg maybe 500pF to 1nF.
A regular guitar with a buffered output sounds very different.

[martin manning]

Your suggestion for a mod looks quite similar to the suggestion posted in my second post?

Yes. Might need to experiment with the mixing resistor values to balance the channel volumes. Should be able to reroute a few things and or lift some component leads to test it.

[ynor]

For the LTP to work properly, the ring of one of the TRS plugs would need grounding.
Or just use a mono TS plug for one, TRS for the other.

Good suggestion, that makes sense: Use a TRS on one of the loops (ex Ch1) for send and receive, and a TS on the other for send only that connects with the tip on the other while grounding the return on that FX-loop (Ch2). Then both loops will be sent to C-lator (or similar) for FXs while only one of the sides of the PI receives, as it would when using only one of the channels. And I guess it doesn't matter which of the sides of the PI is used.

[ynor]

Should be able to reroute a few things and or lift some component leads to test it.

The inside of the Matchless is notoriously difficult to experiment with. There are no turret boards, and the soldering is very difficult to desolder and unwind. I once installed a VVR in my previous DC30 in 2009 and had a really hard time with desoldering things, but then again, I'm an amateur But yes, it could be done, but I prefer an outboard solution.

inside.jpg

inside 2.jpg

[ynor]

double post...

[ynor]

Might need to experiment with the mixing resistor values to balance the channel volumes.

Thought: If building into an external box, potmeters instead of (or in addition to) resistors could act as individual channel volumes..?

[martin manning]

I see, you want to run both loop sends into a mixer, without changing anything upstream and then feed the return into one of the PI inputs. Maybe a pot could be wired as a balance control with the wiper feeding the C-lator’s buffer input? This is hanging more stuff off the signal paths, but would be easy enough to try. Another possibility would be an op-amp mixer stage.